[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

some 6bone registry questions for the list



>>>>> "davidk" == davidk  <[email protected]> writes:

    davidk> Hi,

    davidk> I recently got a few remarks and wanted to know your
    davidk> opinion.

    davidk> The current draft for the registry describes that domain
    davidk> names should be used in the tunnel specification of the
    davidk> 'ipv6-site' objects. This has the advantage that it is
    davidk> very easy to derive the IPv4 *and* IPv6 address through
    davidk> DNS.

    davidk> I have changed the syntax checking recently in such way
    davidk> that it doesn't accept IPv4 numbers anymore but that it
    davidk> will help a bit by doing a reverse lookup. However, if
    davidk> nothing is found, a error is generated.

Please don't do that. I don't think there is any reason why you should
require tunnel end-points to have a DNS address. Remember tunnel end-points
are supposed to be routers... routers don't always DNS records for all of
it's addresses and even if they do, most of the times the IP address is
much more significant as it is unique while the name may point at several
addresses.

    davidk> Guy Davies told me that this might have gone a bit too
    davidk> far. Do you agree?  Should I change it to a warning only
    davidk> or leave it as is with the strong syntax checking?

I personally think you should just accept IP addresses... they are just
much more reliable and meaningful to debug network problems.

    davidk> In another question, it was pointed out to me that RIPng
    davidk> was the better protocol name instead of RIPv6.

Yap... RIPng is RIPv2 with support for IPv6 addresses if i'm not mistaken.
So the name RIPv6 is, IMHO, misleading.

  Pedro.